Skip to main content

Commodity Computing

I have been mulling around with this idea: what if people didn't have to buy powerful computers but instead just buy a computer good enough that connects to a powerful time sharing system which can pretty much be located somewhere else in the world? Will the concept fly? What changes will it bring to the PC consumer industry? What will it mean to software vendors who target the end-user market?

First off, let's consider the proposition: what if you can have your data hosted in a data center which has all the failover and redundancy technology you can only dream about, with bandwidth that you can only wish you could have, and with the most powerful computers you can only wish will fit in your pocket. Now what if you can do that for a recurring cost like how you pay for your electricity, water, bandwidth?

Personal computers are getting cheaper and cheaper, and it would certainly be very easy for some well-off companies/individuals to set up a cluster of computers from these personal computers to come up with a system that shares resources and serves hundreds and thousands of users simultaneously. Will this model work for third world countries like the Philippines?

I think it would. If PLDT sold a computer worth Php 15,000 and gave a low income family a phone line along with it and a year's worth of access to a time-sharing system in their data center. If Php 15,000 is too high even for small income families then let's say that's spread out over a year (roughly Php 1,000++ monthly). For a family with children studying in shool, what they'll just probably need is access to some internet resources through a browser (that ran on the server), a word processor, some educational games, etc. and all this can be accessed though a dialup link via a VNC client.

Now think of the economics on the side of PLDT (or Globe, or Bayantel, or some other telco with internet data centers): A PC with two dual core processors can easily host (4 users per processor, 4 virtual systems) 16 people _easy_. How much does a PC like that cost? Let's check: easily around Php 100k. But with 16 people paying 10k each, that already covers the cost of the actual PC. The internet bandwidth is realy just constant.

How now does PLDT or other telco's make more money out of it? The same way ABS-CBN, GMA, and other companies make money. Only this time, PLDT has direct access to a demographic that these media companies only want to be able to contact directly. Ads drive the entertainment industry, and ads drive the Internet business (more or less). And at the same time, they can do more direct marketing to the people that actually avail of their services. It also allows them to move into a different market segment and re-vitalize their fixed-line business and synergize with the internet business.

This can allow faster Internet adoption in the many different segments of the Philippine market: more than 50% of which cannot afford the luxuries of broadband Internet access.

Now if I only had the capital...

CHill...

Comments

Popular posts from this blog

Writing Again

It's 2019 and I just realised that I've not written on this blog for a long while. I feel a little bad about this so I'm picking it back up again. More importantly, I've limited my social media to just Twitter (I've deleted all my Facebook-related accounts) and will be writing more on the blog instead of engaging in other social media sites. If you want to reach me directly, you can also reach me through my keybase.io account for encrypted communication. If you have my phone number, you can also contact me through Signal. Quite a number of things have happened in the past few years and here's a quick update on things that I can share:

I've been working on XRay, a function call tracing system now part of the LLVM project. This took a good two and some years of my time at Google.Most recently I've moved to the Chrome Operations Team still here in Google Sydney. I can't give specifics yet of what I'll be working on, so stay tuned.There've been c…

From FOMO to JOMO

Until very recently I believed that I needed to be on top of the latest news and happenings not only in my field (computer science and software engineering) but also in as many things as I can be on top of. This meant subscribing to all sorts of magazines, newsletters, YouTube channels, Twitch streamers, watching TV and live sport events, etc. — I was on top of a lot of the latest happenings, trends, news, interesting developments. I was having fun and I felt busy. What I did not feel was particularly effective nor productive. I felt like I was consuming so much information with the thought that it might be useful someday. When I was younger this wouldn’t have been an issue but I realised that ever since I’ve started taking stock of what I’ve been spending my time on, that a lot of it I’ve been spending just staying on top of things that I really didn’t need to be on top of. This article is about some of the realisations I’ve made in the course of exploring this issue of “FOMO” or th…

A Passion Project

I was so moved today by the prospect of a passion project that I took some time on a Friday night to get it done. Let me present the #RedJeans project over at redjeans.org. I've found myself wanting to work on a project that came purely from the heart and one that was very dear to me, something that is personal, and connects with a larger community of people in the world.
The idea for redjeans.org came to me as a hint when I was writing up my reflection for 2018. I realised that I didn't spend quite as much time identifying with and working with a community. I did a bit of soul-searching and found that one of the activities I really enjoyed and cherished in years past is donating blood -- and I keep wondering why not more people do it. It was an idle thought but then a conversation with someone where I described why I wrote down "donate blood more often" in 2019 became an idea where instead of just me doing it, how about if I get my friends to do it too?

I left it a…